Federated learning (FL) is a distributed machine learning paradigm ensuring data privacy. However, the statistical heterogeneity poses a challenge to building a single model that can perform well across all clients’ data distributions. Personalized FL (PFL) has emerged as a solution to mitigate the impact of statistical heterogeneity by training separate models for various target data distributions. A crucial aspect of PFL is to utilize additional information in the federated system to assist in training personalized models. In this article, we introduce a novel PFL approach that leverages a multiview geometry structure (GPFL). GPFL formulates an optimization problem to determine the correlation weights among clients by utilizing the geometry structure composed of client gradient updates. It further builds information carriers that facilitate personalized training based on the weights. To accurately capture the correlations, we use both L2 distance and cosine similarity views to depict geometric similarity. In federated training, the discrepancy in timeliness and tendency to overfit to local data in gradient updates cause the geometric similarity of such updates to inadequately reflect the client relationships. Therefore, GPFL employs representative gradients extracted from the client’s historical gradients to infer the correlation weights. Experimental results on four data sets, convex and nonconvex objectives, and two FL settings demonstrate that our method outperforms several PFL methods.

In recent years, significant progress has been made in detecting, localizing, and classifying multiple magnetic dipole sources (DLCMMS) based on magnetic gradient tensors. Traditional DLCMMS methods face challenges due to limited resolution and high-dimensional parameter optimization. In response, we propose a novel DLCMMS method within the Cooperative Coordination (CC) framework. Specifically, we introduce a decomposition strategy within the CC framework to partition the original DLCMMS problem into lower-dimensional sub-problems. We perform joint parameter optimization for source number, positions, and magnetic moment parameters in each sub-problem covering a subset of magnetic sources and measurement points. Moreover, this letter enhances the hybrid optimization algorithm, combined with differential evolution and Levenberg-Marquardt, for joint parameter optimization. In simulations, the number recognition accuracy of sources is 99.3% without noise, about three times higher than traditional methods. During a field demonstration, the position estimation error was below 0.053m.

Federated learning (FL) is a distributed machine learning framework that allows resource-constrained clients to train a global model jointly without compromising data privacy. Although FL is widely adopted, high degrees of systems and statistical heterogeneity are still two main challenges, which leads to potential divergence and nonconvergence. Clustered FL handles the problem of statistical heterogeneity straightly by discovering the geometric structure of clients with various data generation distributions and getting multiple global models. The number of clusters contains prior knowledge about the clustering structure and has a significant impact on the performance of clustered FL methods. Existing clustered FL methods are inadequate for adaptively inferring the optimal number of clusters in environments with high systems' heterogeneity. To address this issue, we propose an iterative clustered FL (ICFL) framework in which the server dynamically discovers the clustering structure by successively performing incremental clustering and clustering in one iteration. We focus on the average connectivity within each cluster and give incremental clustering and clustering methods that are compatible with ICFL based on mathematical analysis. We evaluate ICFL in experiments on high degrees of systems and statistical heterogeneity, multiple datasets, and convex and nonconvex objectives. Experimental results verify our theoretical analysis and show that ICFL outperforms several clustered FL baseline methods.

Abstract On-line gambling is the key illegal behaviour of public security department in most countries due to the potential threat to cyberspace security and social stability. Hence, the research on gambling domain names (GDN) classification is quite important and in great demand for academia and industry. Till now, there is very little research work on this topic. Most of the GDN training datasets in previous work were chosen from GDN blacklists provided by publicly available data sources, and the authors did not verify the authenticity and accuracy of these datasets, and the classification results are not particularly satisfactory. In this paper, certificated and textual analysis-based classification method CT-GDNC is proposed to get GDN training data set with an accuracy of 0.9776 and significantly improve the classification results of GDN. The exhaustive comparative experiments on 10K GDN obtained via Bert fine-tuning model and 10K benign data collected from Alex Top 1 million list show that the proposed method achieves new baseline result for GDN classification with classification accuracy 0.9936, precision 0.9936, F1 0.9936 and recall 0.9939.

Federated Learning (FL) offers a novel distributed machine learning context whereby a global model is collaboratively learned through edge devices without violating data privacy. However, intrinsic data heterogeneity in the federated network can induce model heterogeneity, thus posing a great challenge to the server-side model aggregation performance. Existing FL algorithms widely adopt model-wise weighted averaging for client models to generate the new global model, which emphasizes the importance of the holistic model but ignores the importance of distinctions between internal parameters of various client models. In this paper, we propose a novel parameter-wise elastic weighted averaging aggregation approach to realize the rapid fusion of heterogeneous client models. Specifically, each client evaluates the importance of model internal parameters in the model update and obtains the corresponding parameter importance coefficient vector; the server implements the parameter-wise weighted averaging for each parameter based on their importance coefficient vectors, thereby aggregating a new global model. Extensive experiments on MNIST and CIFAR-10 datasets with diverse network architectures and hyper-parameter combinations show that our proposed algorithm outperforms the existing state-of-the-art FL algorithms on the performance of heterogeneous model fusion.

The Tolley-Lawson(TL) model plays an important role in the aeromagnetic survey. The TLG model based on the TL model introduced the correction of the geomagnetic gradient in the real world and improved the precision of aeromagnetic compensation. The key to geomagnetic gradient compensation is to introduce the system’s position information. However, in general, there is an offset between the origin of the coordinates of the magnetometer and the fuselage, which will lead to sensor position errors, especially during maneuvers. Correcting the positional error will eliminate the offset and increase the precision of the aeromagnetic survey. In this paper, the principle of the positional error is analyzed and an improved model(TLG-C) aims to correct the positional error is proposed. The positional error is modeled as a function of the attitude angle of the aircraft and the distance between the fuselage and magnetometer. An improved aeromagnetic compensation algorithm is derived based on the TLG-C model. The results of 24 cases in actual calibration flights show that the proposed method can effectively enhance the improvement ratio.

The estimating of the Tolles-Lawson model's coefficients plays an important role in aeromagnetic compensation. The directional cosines, which are indispensable for estimating the coefficients, are usually measured by a three-axis strapdown magnetometer in a general aeromagnetic survey system. However, in some cases, the scalar magnetometer may have a much higher accuracy than the three-axis strapdown magnetometer. This imbalance of the measurement accuracy is then introduced into the Tolles-Lawson model and affects the estimation of the coefficients. In this letter, a modified Tolles-Lawson model is introduced to reduce the imbalance through substituting the error model of the three-axis strapdown magnetometer into the calculation of the directional cosines. The characteristics of the modified model are analyzed and the corresponding coefficient-estimating system is developed. Simulation results illustrate that the modified model is more robust to the larger errors of the three-axis strapdown magnetometer than the classical model.

Pseudorandom sequences are very important in the field of cryptography. The characteristics such as non‐linearity and random‐like behaviours make chaotic systems suited to generate pseudorandom sequences. However, most of chaos‐based pseudorandom number generators have a typical shortcoming. That is, the finite precision in all processors may cause the chaotic systems to degenerate into a periodic function or a fixed point. To overcome this shortcoming, a hyperchaos‐based generator is proposed. First, a new hyperchaotic system with bigger Lyapunov exponent is constructed. Then the self‐shrinking generator, which is superior to many other linear feedback shift register‐based generators, is used to perturb the hyperchaotic sequences to decrease the period degeneration and improve the performance of the sequences. The proposed generator is named as hyperchaos with self‐shrinking perturbance generator (H‐SSP generator). The analysis results show that the H‐SSP generator has better performance.